Your SaaS Churn Is 8%. Is That a Death Spiral or a Fixable Gap?

First, Confirm Whether 8% Is Monthly or Annual

An 8% churn figure is not one problem. If that number is monthly and it holds for a year, you would lose about 63% of the starting customer base, because monthly rates compound: 1 − (1 − 0.08)¹². That is a leak sales cannot outrun for long. If the same 8% is annual, it is above the current SaaS medians discussed below, but it is still in a range where a focused retention plan can close the gap.

Before you treat 8% as a death spiral, write down the period, the formula, and whether you are counting customers or revenue. Those three choices change the diagnosis more than any single "industry average."

What Current Benchmarks Actually Say

Recurly's July 2026 churn benchmarks, drawn from its subscription network, are labeled as annual figures. For SaaS, the page gives a median of 3.22% (2.16% voluntary and 1.06% involuntary). In another passage it quotes 3.04% for software businesses, with top-quartile performers at 1.78% or below. Recurly describes below 2% annual churn as strong, 2% to 4% as the range where most well-run subscription businesses operate, and above 5% as worth investigating.

An earlier version of that same Recurly page, archived in January 2026 and based on 2023 data, showed an overall rate of 3.27% and said those churn rates were calculated monthly. That older wording helps explain why a figure near 3.5% still circulates as a monthly target. Recurly's current page should be read as annual, so comparing a monthly 8% with 3.22% mixes two different clocks.

Other datasets are much higher because they measure a different population. Churnkey's November 2025 analysis of Stripe data puts SaaS annual customer loss at 38%, and says involuntary churn accounts for 22% of that SaaS total. Use Recurly to compare subscriber churn on a billing network; use the Stripe/Churnkey figures to remember that some panels of smaller or more transactional businesses look very different. Do not average them.

Why a Single Benchmark Misleads

Even inside Recurly's own tables, the "right" comparison depends on who you sell to. Education has the highest median in the July 2026 industry table, at 4.99% annual (3.30% voluntary and 1.69% involuntary). By average revenue per customer, median annual churn is 4.29% in the $10 to $25 band and 2.87% in the $100 to $250 band. Recurly also notes that in the $250+ ARPC cohort, median annual churn is 3.54%, but involuntary churn falls to 0.18%.

The practical rule is simple: compare your rate with peers on period, contract value, and motion, not with a blended number you heard in a meeting. If early-tenure accounts are driving the losses, the work may start in onboarding rather than in late-stage save plays. Our guide to why customers churn after a smooth onboarding covers that post-go-live gap.

Logo Churn vs. Revenue Churn

Eight percent of customers leaving is not the same as eight percent of ARR leaving. Logo churn can look ugly while net revenue still holds up, if small accounts cancel and larger ones expand. Revenue churn can look ugly while logo churn stays modest, if one large contract downgrades or leaves. For the revenue math, see how to calculate and improve net revenue retention.

Voluntary vs. Involuntary Churn

Voluntary churn is a decision: the customer did not see enough value, did not adopt the product, or chose a competitor. Involuntary churn is usually a payment failure: an expired card, a declined renewal, or a billing error. Recurly's SaaS split, 2.16% voluntary and 1.06% involuntary, means payment failure is about one-third of that median. Churnkey's Stripe-based SaaS table puts the involuntary share at 22% of a much higher annual loss rate.

Those are different panels, but they agree on the operating point: do not spend a product quarter on a problem that is actually in dunning and account-updater work, and do not treat a failed-payment queue as proof that the product missed the mark. Split the 8% before you assign owners.

Leading Indicators Beat a Cancellation Email

A cancel request is late in the story. Usage that shrinks, features that fall out of the weekly workflow, and support threads that turn sharp usually show up earlier. There is no universal "90-day" clock that fits every contract, but the useful habit is the same: compare the recent usage, adoption, and support pattern of accounts that left with the pattern of accounts that stayed, then watch current accounts for the same turn.

Our guides to finding why customers leave and to customers who are already slipping without the team seeing it go deeper on that diagnosis. If the losses cluster right after implementation, start with why a customer can still churn after onboarding went well.

Health Scores Only Help When They Are Segmented

An unsegmented 8% hides the playbook. Month-one logos that never reach first value need a different response than four-year accounts being replaced. Build health around the cohorts you actually manage: tenure, ACV, product, and channel.

Quivly is built for that kind of read, not as a billing-recovery tool. According to How Quivly Works, it unifies CRM, billing, support, calls, Slack, Gmail, and product usage into one customer profile. Health scores assess each customer across revenue, product usage, engagement, support, and market signals, with configurable category weights. Agents can run when a health bucket changes. Plays investigate outcomes such as stalled adoption or renewal risk, and the Actions inbox is where recommended outreach can be reviewed before it is sent.

That does not replace a payment-recovery program, and it does not guarantee a fixed lead time before cancellation. It does give the team one place to see whether the 8% is a usage problem, a support problem, a revenue problem, or a quiet relationship, and to act with evidence.

What to Do Once You Know Which 8% You Have

1. Name the metric. Period, logo versus revenue, and voluntary versus involuntary, written down for the same date range.
2. Recover involuntary churn as an operations problem. Retry logic, account updater services, and clear failed-payment outreach belong here, independent of the product roadmap.
3. Fix early voluntary churn at the adoption gap. Time to first value, a missed core workflow, and silence after go-live are the usual tells.
4. Trigger save work from a segmented health drop, not from a single blended rate. Put context and a next step in front of the owner, then review anything that goes to the customer.
5. Use expansion to put logo losses in financial context. NRR tells you whether the remaining base is covering what left. It is not a substitute for stopping a real leak.

Conclusion

Eight percent is a prompt to diagnose, not a verdict. If it is monthly, the compounding math is severe and the work is urgent. If it is annual, Recurly's current SaaS median of 3.22% says you are above the typical range, but not past the point where payment recovery, early adoption, and segmented save work can still change the result. Match the benchmark to your period and your customers, then fix the part of the 8% you can actually move.

If you'd like to see how Quivly would score risk and queue the next step across your book, book a demo with Quivly.

Sources

1. Churn Rate Benchmarks | Recurly (July 2026 data) - recurly.com
2. Customer churn benchmarks | Recurly (archived January 2, 2026) - web.archive.org
3. Involuntary Churn Benchmarks and Tactical Advice | Churnkey (November 14, 2025) - churnkey.co
4. How Quivly Works | Quivly Docs - docs.quivly.ai
5. Health Scores | Quivly Docs - docs.quivly.ai
6. Agents | Quivly Docs - docs.quivly.ai
7. Plays | Quivly Docs - docs.quivly.ai
8. Actions | Quivly Docs - docs.quivly.ai